The test did not wait for the mount command to fully shutdown all
running goroutines. This caused the go race detector to report a data
race related to lock refreshes.
==================
WARNING: DATA RACE
Write at 0x0000021bdfdb by goroutine 667:
github.com/restic/restic/internal/backend/retry.TestFastRetries()
/restic/restic/internal/backend/retry/testing.go:7 +0x18f
github.com/restic/restic/cmd/restic.withTestEnvironment()
/restic/restic/cmd/restic/integration_helpers_test.go:175 +0x183
github.com/restic/restic/cmd/restic.TestMountSameTimestamps()
/restic/restic/cmd/restic/integration_fuse_test.go:202 +0xac
testing.tRunner()
/usr/lib/go/src/testing/testing.go:1446 +0x216
testing.(*T).Run.func1()
/usr/lib/go/src/testing/testing.go:1493 +0x47
Previous read at 0x0000021bdfdb by goroutine 609:
github.com/restic/restic/internal/backend/retry.(*Backend).retry()
/restic/restic/internal/backend/retry/backend_retry.go:72 +0x9e
github.com/restic/restic/internal/backend/retry.(*Backend).Remove()
/restic/restic/internal/backend/retry/backend_retry.go:149 +0x17d
github.com/restic/restic/internal/cache.(*Backend).Remove()
/restic/restic/internal/cache/backend.go:38 +0x11d
github.com/restic/restic/internal/restic.(*Lock).Unlock()
/restic/restic/internal/restic/lock.go:190 +0x249
github.com/restic/restic/cmd/restic.refreshLocks.func1()
/restic/restic/cmd/restic/lock.go:86 +0xae
runtime.deferreturn()
/usr/lib/go/src/runtime/panic.go:476 +0x32
github.com/restic/restic/cmd/restic.lockRepository.func2()
/restic/restic/cmd/restic/lock.go:61 +0x71
[...]
Goroutine 609 (finished) created at:
github.com/restic/restic/cmd/restic.lockRepository()
/restic/restic/cmd/restic/lock.go:61 +0x488
github.com/restic/restic/cmd/restic.lockRepo()
/restic/restic/cmd/restic/lock.go:25 +0x219
github.com/restic/restic/cmd/restic.runMount()
/restic/restic/cmd/restic/cmd_mount.go:126 +0x1f8
github.com/restic/restic/cmd/restic.testRunMount()
/restic/restic/cmd/restic/integration_fuse_test.go:61 +0x1ce
github.com/restic/restic/cmd/restic.checkSnapshots.func1()
/restic/restic/cmd/restic/integration_fuse_test.go:90 +0x124
==================
Counting the first occurrence of a duplicate blob as used and counting
all other as duplicates, independent of which instance of the blob is
kept, is only accurate if all copies of the blob have the same size. This
is no longer the case for a repository containing both compressed and
uncompressed blobs.
Thus for duplicated blobs first count all instances as duplicates and
then subtract the actually used instance later on.
As long as only a small fraction of the data in a repository is
rewritten, the keepBlobs set will be rather small after cleaning it up.
As golang maps do not shrink their memory usage, just copy the contents
over to a new map. However, only copy the map if the cleanup removed at
least half the entries.
The set covers necessary, existing and duplicate blobs. This removes the
duplicate sets used to track whether all necessary blobs also exist.
This reduces the memory usage of prune by about 20-30%.
The RetryBackend tests depend on the mock backend. When the Backend
interface is eventually split from the restic package, this will lead to
a dependency cycle between backend and backend/mock. Thus split the
RetryBackend into a separate package to avoid this problem.
The comparison of the current time and the last lock refresh were using
seconds represented as integers. As the test only waits for up to one
second, the associated number truncation can cause the test to take
longer than once second and thus to fail.
Switch to nanoseconds to avoid this problem. This also slightly speeds
up the test.
FindFilteredSnapshots no longer prints errors during snapshot loading on
stderr, but instead passes the error to the callback to allow the caller
to decide on what to do.
In addition, it moves the logic to handle an explicit snapshot list from
the main package to restic.
The only use cases in the code were in errors.IsFatal, backend/b2,
which needs a workaround, and backend.ParseLayout. The last of these
requires all backends to implement error unwrapping in IsNotExist.
All backends except gs already did that.
Monotonic timers are paused during standby. Thus these timers won't fire
after waking up. Fall back to periodic polling to detect too large clock
jumps. See https://github.com/golang/go/issues/35012 for a discussion of
go timers during standby.
Restic continued e.g. a backup task even when it failed to renew the
lock or failed to do so in time. For example if a backup client enters
standby during the backup this can allow other operations like `prune`
to run in the meantime (after calling `unlock`). After leaving standby
the backup client will continue its backup and upload indexes which
refer pack files that were removed in the meantime.
This commit introduces a goroutine explicitly monitoring for locks that
are not refreshed in time. To simplify the implementation there's now a
separate goroutine to refresh the lock and monitor for timeouts for each
lock. The monitoring goroutine would now cause the backup to fail as the
client has lost it's lock in the meantime.
The lock refresh goroutines are bound to the context used to lock the
repository initially. The context returned by `lockRepo` is also
cancelled when any of the goroutines exits. This ensures that the
context is cancelled whenever for any reason the lock is no longer
refreshed.
Previously the global context was either accessed via gopts.ctx,
stored in a local variable and then used within that function or
sometimes both. This makes it very hard to follow which ctx or a wrapped
version of it reaches which method.
Thus just drop the context from the globalOptions struct and pass it
explicitly to every command line handler method.
We can either preallocate storage for a file or sparsify it. This
detects a pack file as sparse if it contains an all zero block or
consists of only one block. As the file sparsification is just an
approximation, hide it behind a `--sparse` parameter.
`restic unlock` now only shows `successfully removed locks` if there were locks to be removed.
In addition, it also reports the number of the removed lock files.
Sending data through a channel at very high frequency is extremely
inefficient. Thus use simple callbacks instead of channels.
> name old time/op new time/op delta
> MasterIndexEach-16 6.68s ±24% 0.96s ± 2% -85.64% (p=0.008 n=5+5)
This results in printing a `(default: $ENV) (default: value)` suffix for
the corresponding options which looks strange. In addition, some of the
environment variables might contain secrets which should not be
displayed.
`init` and `copy` use `--repo2` with two different meaning which has
proven to be confusing for users. `--from-repo` now consistently marks a
source repository from which data is read. `--repo` is now always the
target/destination repository.
After repacking every blob that should be kept must have been repacked.
We have seen a few cases in which a single blob went missing, which
could have been caused by a bitflip somewhere. This sanity check might
help catch some of these cases.
Unused blobs are not a problem but rather expected to exist now that
prune by default does not remove every unused blob. However, the option
has caused questions from users whether a repository is damaged or not,
so just remove that option.
Note that the remaining code is left intact as it is still useful for
our test cases.
Use runtime.GOMAXPROCS(0) as worker count for CPU-bound tasks,
repo.Connections() for IO-bound task and a combination if a task can be
both. Streaming packs is treated as IO-bound as adding more worker
cannot provide a speedup.
Typical IO-bound tasks are download / uploading / deleting files.
Decoding / Encoding / Verifying are usually CPU-bound. Several tasks are
a combination of both, e.g. for combined download and decode functions.
In the latter case add both limits together. As the backends have their
own concurrency limits restic still won't download more than
repo.Connections() files in parallel, but the additional workers can
decode already downloaded data in parallel.
Previously, SaveAndEncrypt would assemble blobs into packs and either
return immediately if the pack is not yet full or upload the pack file
otherwise. The upload will block the current goroutine until it
finishes.
Now, the upload is done using separate goroutines. This requires changes
to the error handling. As uploads are no longer tied to a SaveAndEncrypt
call, failed uploads are signaled using an errgroup.
To count the uploaded amount of data, the pack header overhead is no
longer returned by `packer.Finalize` but rather by
`packer.HeaderOverhead`. This helper method is necessary to continue
returning the pack header overhead directly to the responsible call to
`repository.SaveBlob`. Without the method this would not be possible,
as packs are finalized asynchronously.
raw-data summed up the size of the blob plaintexts. However, with
compression this makes little sense as the storage size in the
repository is lower due to compression. Thus sum up the actual size each
blob takes in the repository.
The GlobalOptions struct now embeds a backend.TransportOptions, so it
doesn't need to construct one in open and create. The upload and
download limits are similarly now a struct in internal/limiter that is
embedded in GlobalOptions.
There were three loops over the index in restic prune, to find
duplicates, to determine sizes (in pack.Size) and to generate packInfos.
These three are now one loop. This way, prune doesn't need to construct
a set of duplicate blobs, pack.Size doesn't need to contain special
logic for prune's use case (the onlyHdr argument) and pack.Size doesn't
need to construct a map only to have it immediately transformed into a
different map.
Some quick testing on a 160GiB local repo doesn't show running time or
memory use of restic prune --dry-run changing significantly.
github.com/pkg/errors is no longer getting updates, because Go 1.13
went with the more flexible errors.{As,Is} function. Use those instead:
errors from pkg/errors already support the Unwrap interface used by 1.13
error handling. Also:
* check for io.EOF with a straight ==. That value should not be wrapped,
and the chunker (whose error is checked in the cases changed) does not
wrap it.
* Give custom Error methods pointer receivers, so there's no ambiguity
when type-switching since the value type will no longer implement error.
* Make restic.ErrAlreadyLocked private, and rename it to
alreadyLockedError to match the stdlib convention that error type
names end in Error.
* Same with rest.ErrIsNotExist => rest.notExistError.
* Make s3.Backend.IsAccessDenied a private function.
Tree packs are cached locally at clients and thus benefit a lot from
being compressed. Ensure this be having prune always repack pack files
containing uncompressed trees.
The `stats` command checks inodes to not count hardlinked files multiple
times into the restore size. This check applies across all snapshots and
not only within snapshots. As a result the result size was far too low
when calculating it for multiple snapshots and it would vary depending
on the order in which snapshots were listed.
The new option allows prune to operate with nearly no scratch space by only removing
no longer necessary pack files and first deleting the index before
rebuilding it. By first deleting the index it becomes safe to just
delete no longer necessary pack files. However, as a downside there's
now the risk that the repository becomes inaccessible if prune fails.
To recover from that problem a user might have to manually delete the
repository index and then run (a full) `rebuild-index` again.
A compressed index is only about one third the size of an uncompressed
one. Thus increase the number of entries in an index to avoid cluttering
the repository with small indexes.
As an exception prune is still allowed to load the index before
snapshots, as it uses exclusive locks. In case of problems with locking
it is also better to load snapshots created after loading the index, as
this will lead to a prune sanity check failure instead of a broken snapshot.
When resolving snapshotIDs in FindFilteredSnapshots either
FindLatestSnapshot or FindSnapshot is called. Both operations issue a
list operation to the backend. When for example passing a long list of
snapshot ids to `forget` this could lead to a large number of list
operations.
These commands filter the snapshots according to some criteria which
essentially requires loading the index before filtering the snapshots.
Thus create a copy of the snapshots list beforehand and use it later on.
During a backup the index is written before the corresponding snapshots.
To ensure that a concurrent/later restic run can read a snapshot's data,
restic thus must first load the snapshots and only afterwards the index.
Otherwise it is not possible to ensure that the loaded index is recent
enough to cover all of the snapshot's data.
